Modeling The Unemployment Rate At The Eu Level By Using Box-Jenkins Methodology

Abstract

Unemployment, as a measure of market conditions, appears as a crucial economic problem and a phenomenon with considerable negative social consequences, and, as such, requires attention and adequate approach to finding solutions. Enormous unemployment rates are a reality not only in developing and transition countries, but also in some developed countries. Inadequately conducted privatization, unsuccessful transfer of workers from the public to the private sector, inefficiency in attracting foreign direct investment, and the world economic crisis of 2008 have made unemployment a universal disease of modern society. The paper presents economic models in which the unemployment rate is the central analyzed phenomenon. In this context, an important task of European economic policy-makers is to project future unemployment rates. Box-Jenkins methodology, i.e. the seasonal ARIMA model, is one approach to the modeling of time series, or, more specifically, for forecasting future values. The subject of this paper is the analysis of the evolution of the unemployment problem on the basis of the values in the period from 2000 to 2015, based on the case of 28 countries of the European Union. Building on the research subject, the purpose of the paper is to create the statistical model for forecasting the values of the monthly unemployment rates in the European Union for the future and establishing its trend.

Comparative Analysis of Different Univariate Forecasting Methods in Modelling and Predicting the Romanian Unemployment Rate for the Period 2021–2022

“…In the European Union, the unemployment rate is forecasted using Box-Jenkins and TRAMO/SEATS methods [36,37]. In European countries, the unemployment rate was predicted using the Box-Jenkins methodology in Germany using the ARIMA and VAR models [38], in the Czech Republic using the SARIMA model [39,40], in the German regions using a model spatial GVAR [41], in Greece, both as a dynamic process and as a static process using SARIMA models [42,43], and in Slovakia using ARIMA and GARCH models [44].…”